Response surface method and experimental design were applied as alternatives to the conventional methods for optimization of the coagulation test. A central composite design was used to build models for predicting and optimizing the coagulation process. The model equations were derived using the least square method of the Minitab 16 software. In these equations, the removal efficiency of turbidity and COD were expressed as second-order functions of the coagulant dosage and coagulation pH. By applying RSM, the optimum condition using PFPD1 was coagulant dosage of 384 mg/L and coagulation pH of 7.75. The optimum condition using PFPD2 was coagulant dosage of 390 mg/L and coagulation pH of 7.48. Confirmation experiment demonstrated a good agreement between experimental values and model predicted. This demonstrates that RSM and CCD can be successfully applied for modeling and optimizing the coagulation process using PFPD1 and PFPD2.